The Pricing Problem in Health Care

Part of the problem we have with medical care—in the U.S. and elsewhere—is that it is incredibly difficult to price. That would be true in any circumstance, but it's especially true given the lack of competition in the health care market. Reason contributor Arnold Kling has been thinking about this problem, and he has a great post looking at the pricing problem in the context of the Medicare payment system.

For both governments and private payers and providers, it's very, very difficult to figure out how to value health care: Do you price based on input—the work (including education) that went into the service? Based on the value of the outcome to the individual? The value to society of the medical benefit? Medicare's fee-for-service system ends up paying doctors based on quantity rather than quality. It's not a true universal payer, like some other countries have set up, but the pay rates it uses exhibit a significant influence on the rest of the market. And even outside the influence of Medicare payment rates, private insurers don't have a clear way to calculate rates. Kling argues that a system of competitive HMOs, each allowed to experiment with different provider groups and payment methods, might help solve the problem:

The socialist calculation problem is something that I worry about a great deal in the case of medical services. The insurance company is the socialist planner in this case. How does it know how much the consumer values a given medical service at the margin? How does it know the opportunity cost of the medical service?

I can see the problem being solved more easily in the context of a competitive market in HMO's. Each HMO guesses what combination of medical services at what price will maximize profits, in the same way that an automobile company guesses what combination of features at what price will maximize profits. No one solves the problem perfectly, but they keep groping for better answers.

In my mind, each HMO can produce its services using whatever combination of inputs it chooses. It can have physicians trained in medical schools, or it can train doctors internally. All forms of practice regulation and licensing have disappeared, and HMO's compete in part on the basis of reputation. Some medical services will be bungled, but that happens a lot in today's system, also.

The key here is allowing widespread experimentation. If you do that, some pricing and service models will fail. Some will succeed a little. A few might succeed a lot. But allowing experimentation in the health care sector, especially on the business side, which tends to control access and manage resource allocation, is tough in part because people are (understandably) afraid of failure.

In most markets, if a system fails, people move on. But in the health care market, failure—or even just the variety of outcomes that come from widespread experimentation—is far more fraught. We want everyone to have equal access and equal care, and that care should be the best possible care. Varied outcomes aren't acceptable. And so we set up complex webs of regulations and licensing that attempt to guarantee that quality care is available to everyone. But the result is that we get stuck with business-side, resource-related problems—like, say, optimal pricing and service bundling—and don't get to experiment with ways to fix them. So progress proceeds at a halting, bureaucratic pace, a project of careful, politically sensitive technocrats rather than entrepreneurs. We get state-run pilot programs. And payment system reforms. And formulas and bureaucracies charged with holding cost growth in check. Maybe some progress is made at the margins, but for the most part, the resource-allocation problems grow. And all the experts sit around wondering how to fix those problems, but afraid to take the one step that's most likely to help, which is to let entrepreneurs innovate and experiment—and yes, to fail, sometimes—until they find some arrangement that works better.